Abstract

                                Weather Research and Forecasting (WRF) model is a numerical weather prediction system and having wide range of applications.It offers multiple physics options that could be combined in any way to improve the accuracy of forecast suitable to particular region.Each physics option has it’s own individuality and no single scheme could perform better in all locations. In this context, a study was carried out to identify the best WRF microphysics scheme for improvingmedium range rainfallforecastissued by Tami Nadu Agricultural University at block levelforTamil Nadu. Based on the review of previous studies, the present study was restricted to three microphysics options which are suitable for tropical conditions viz. Kessler scheme, WSM3 and WSM6 class schemes. Forecast wasdeveloped daily at 3 km resolution with a lead time of six days for 32 locations of Tamil Nadu and for the first fortnight of November 2017. The simulated forecast values were verified with the observed data collected from the Automatic Weather Stations of TamilNadu Agricultural Weather Network (TAWN) using skill score and root mean square error verification methods. Input data resolution and microphysics option had huge impact on model performance. Data resolution of 0.25 degree performed better than 0.50 degree in all microphysics schemes considered for the study. Among the three microphysics options, the Kessler scheme had higher rainfall forecast accuracy than WSM 3 class scheme and WSM 6 class scheme. Root mean square error was also low in Kessler scheme compared to other two schemes. Kessler scheme produced highly usable forecast than WSM 3 and 6 class schemes. Hence, it is concluded that the Kessler scheme with 0.25o input data resolution produced higher rainfall forecast accuracy for Tamil Nadu region

Key words : :Weatherforecast, WRF, Microphysics,Accuracy, Verification